**1. Introduction**

Prediction models are widely used in the clinic to estimate the risk (or probability) of existing disease or outcome for an individual, determined by the possible values of one or more predictors. In the case of patients with azoospermia due to spermatogenetic dysfunction (also termed non-obstructive azoospermia—NOA), the probability of surgically retrieving sperm from one or both testes represents the outcome that needs to be estimated. Since the ability to predict such an outcome would allow the urologist to individuate those patients who are suited for microdissection testicular sperm extraction (mTESE), several prediction models have been developed to date, however their resulting prediction accuracy was never strong enough to translate their results to the clinical practice. Few candidate predictors have been proposed to be associated with better chances of successful sperm retrieval (SSR), but a consensus has not been reached about them. As a result, actually no clinical or laboratory factor may be used to counsel patients with NOA about their chances of mTESE success.

Indeed, there are some issues that may explain these findings. The most important one is that, in patients with NOA, the testicular parenchyma is not rarely characterized by a highly heterogeneous distribution of histologically and functionally distinct seminiferous tubules (STs), so that the retrieval of sperm is mostly dependent upon the skill and experience of the urologist, his/her learning curve being strictly correlated with the outcome of mTESE [1–3], rather than upon the severity of the spermatogenic dysfunction. In addition, the definition of SSR is not homogeneous among groups: ideally, SSR is defined as the retrieval of an adequate number and quality of sperm for intracytoplasmic sperm injection

**Citation:** Caroppo, E.; Colpi, G.M. Prediction Models for Successful Sperm Retrieval in Patients with Non-Obstructive Azoospermia Undergoing Microdissection Testicular Sperm Extraction: Is There Any Room for Further Studies?. *J. Clin. Med.* **2021**, *10*, 5538. https:// doi.org/10.3390/jcm10235538

Academic Editor: Hiroshi Okada

Received: 7 November 2021 Accepted: 24 November 2021 Published: 26 November 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

(ICSI); however, at least in some cases, the difference between successful (positive outcome) and failed sperm retrieval (negative outcome) may not always be as sharp as it should be to avoid the risk of misclassification.

This notwithstanding, the number of prediction models being evaluated in this field is growing. To establish whether the current knowledge about prediction of mTESE success may justify further studies, in the present article, we will review the evidence about the predictive ability of the clinical and laboratory factors that have been previously proposed as candidate predictors of mTESE outcome.

## **2. Clinical Factors**

Some prediction models of successful sperm retrieval have evaluated the predictive ability of clinical conditions that may be involved in the etiology of NOA (Klinefelter's syndrome, Y chromosome microdeletions, cryptorchidism, varicocele), or may represent putative prognostic factors of mTESE success (testicular volume).

## *2.1. Klinefelter Syndrome*

Klinefelter syndrome (KS) is the most common chromosomal abnormality in men, and is found in about 3–4% of infertile men and in more than 10% of azoospermic men [4]. KS men have typically small, atrophic testes and hypergonadotropic hypogonadism, with tubular hyalinization as the prevalent histopathological pattern. Their genetic profile is characterized in 85–90% of cases by the presence of a supernumerary X chromosome (47, XXY karyotype), while the remaining patients show a mosaic karyotype (46, XY/47, XXY), or rarely, a super-numerous sex chromosome [5]. Despite the severe spermatogenic dysfunction, 8% of patients may have sperm in the ejaculate [6], while testicular sperm may be retrieved in 20–66% of KS men by means of mTESE (see Table 1). Such a wide range of sperm retrieval rates (SRR) may be explained by the unique testicular architecture found in men with KS, who may have sperm in focal enlargements of otherwise sclerotic tubules, instead of having sperm throughout a uniformly dilated tubule [7], so that only a meticulous search within these very small testes may be successful. In addition, as summarized in Table 1, some studies sugges<sup>t</sup> that SRRs may be affected by age (younger patients have better SRRs) or preoperative testosterone level (normal testosterone level is associated with better SRRs).

The predictive role of KS on SSR is still debated. A neural computational model built on 1026 men with NOA demonstrated that KS significantly predicted SSR (OR 3.07 (1.84–5.03), *p* < 0.001) [8]. On the other hand, a meta-analysis evaluating 117 studies enrolling 21,404 patients showed that SSR decreased as a function of the number of KS subjects included in the population of NOA (S = −0.02( −0.04; −0.01); *p* < 0.01) [9]. Still, different surgical methods of sperm retrievals, different surgeons and embryologists' skill and experience, and heterogeneities in patients' characteristics may explain such conflicting results. Further studies should clearly provide information about patients' ages, as well as surgeon's learning curve, to allow the correct interpretation of data.

#### *2.2. Y Chromosome Microdeletions*

The global prevalence of AZF microdeletions in infertile men is estimated to be 7% (95% CL 6.74–6.79) [10]. The most frequently deleted locus in infertile men is AZFc (60– 70%), followed by AZFa (0.5–4%), AZFb (1–5%) and AZFb+c (1–3%) deletion [10]. Men with complete AZFa and AZFb deletions are azoospermic, and sperm cannot be surgically retrieved [11]. A study reported that 3 out of 15 patients with AZFb deletions had sperm on mTESE [12]; however, the Authors defined the AZFb deletions using sY127 and sY134 marker, while classically, the AZFb locus is proximally defined by sY108 and distally characterized by sY134 or sY135; therefore, a partial AZFb deletion could not be excluded in such cases. Men with complete AZFc deletions may have sperm in the ejaculate or be azoospermic, but with good chances of SSR: a recent review reporting the results of 32 studies found that sperm could be retrieved in 13 to 100% of cases, particularly when

mTESE was used [11]. Thus, AZFc deletion may confer better chances of SSR to patients with NOA.
